This study focuses on the morphology formation and morpho-function in a linear-cluster robotic system which is composed of linearly connected autonomous mobile modules. The equations of motion of the linear-cluster robotic system are derived under dynamic interaction with its physical environment, and computer simulations are performed for two cases of physical environment condition: one is for two dimensional friction plane and the other for a maze-like arena surrounded with walls. When each module has simple wIes of locomotion determined according to the direction of interaction force acting from a neighboring module, various morphologies have been formed. When adding intelligence of obstacle avoidance and locomotion towards the object and goal signals to the modules, the linear-cluster robotic system has achieved the task to find the object. enclose it and then carry the object to the goal destination by changing its morphology. Finally some of the future works are discussed from the view of balancing the complexities of morphology and intelligence in cluster robotic systems.